to the editor: We enjoyed the succinct review of cellulitis by Dr. Stulberg and colleagues.¹ As the article emphasized, skin and soft tissue infections are routinely encountered by the primary care physician, and an understanding of the manifestations of bacterial skin infections is crucial for a favorable outcome. However, the diagnostic criteria for cellulitis are very broad and can be considered nebulous.² We have suspected that many of the presentations labeled as cellulitis may represent alternative diagnoses that are often treatable. We were able to verify these suspicions during the performance of a related investigation at our institution.

The infectious disease service at our medical center conducted a trial of duration of therapy for patients with cellulitis. We received referrals from primary care and urgent care clinics throughout our institution, which are staffed by board-certified nurse practitioners, physician assistants, and physicians. Exclusion criteria included cutaneous abscesses requiring drainage, antibiotic use for longer than 24 hours before referral, the presence of diabetic foot ulcers or other chronic nidi of infection, neutropenia, and bite wounds.

During the course of our study, 169 patients were referred with the diagnosis of cellulitis. We observed that 23 patients (13.6 percent) referred for cellulitis actually had an alternative diagnosis. The alternate diagnoses in these patients included the following: abscess requiring incision and drainage (seven patients); abscess not requiring incision and drainage (six patients); herpes zoster (two patients); septic bursitis (two patients); and one patient each with herpetic whitlow (Herpes simplex virus), gangrenous foot, gout, foot fracture, septic arthritis, and tinea pedis. Each patient with an abscess also had overlying cellulitis. Because referral criteria excluded abscesses, we concluded that these deep infections had been undetected. Ten patients required surgical referral for incision and drainage, and most of the conditions required therapy either in addition to or instead of the antibiotics commonly prescribed for cellulitis.

Therefore, we conclude that an additional word of caution should be appended to your recent review of cellulitis, that the diagnosis is sometimes misleading and that alternative conditions exist that closely mimic uncomplicated cutaneous infections. The importance of recognizing and draining purulent collections cannot be overemphasized. Mimicking conditions may not be evident after a good history and examination are performed, and still may not be diagnosed after attempts at aspiration or drainage are completed. The opportunity for effective treatment of these alternate diagnoses may occur only if careful follow-up of the patient is performed.